EDITORIAL

Cryptantigens: time to uncover the real significance

of T-activation

A
s a resident rotating through the transfusion pioneering work in RBC serology, frequently saw polyag-
medicine service at Children’s National Medi- glutination secondary to environmental contamination
cal Center in Washington, DC, many years by bacteria. This phenomenon could also be used to
ago, I had a session with Naomi Luban, about potentiate reactivity in the cold, also demonstrating the
T-activation and its importance for pediatric transfu- first uses of enzyme treatment to aid in antibody identi-
sion. At that time, lectins were readily available com- fication.4-6 Later, Friedenreich discovered that certain
mercially, and the AABB’s Technical Manual had a hidden antigens became exposed due to the actions of
procedure dedicated to their use. I learned the impor- bacterial enzymes, and these hidden or “crypt-” anti-
tance of identifying T-activation in children and then gens were responsible for the unexpected agglutination.
providing the appropriate transfusion therapy. Decades This series of observations was called the Thomsen-
later I became aware of the ongoing controversy sur- Hu € bener-Friedenreich phenomenon and later T-activation,
rounding the presence of T-activation and whether it after Dr Thomsen.7-9 In 1938, the first in vivo case of poly-
had a direct role in red blood cell (RBC) destruction agglutination was described in a 4-year-old with pneumo-
after transfusion. In this issue of TRANSFUSION, Moh- coccus. The etiology was felt to be a soluble substance
Klaren and colleagues1 present a case of an infant with from the bacteria or medications used to treat the patient’s
necrotizing enterocolitis (NEC) and T-activation who infection.3 A variety of bacteria and viruses have been
died in temporal relationship to receiving blood products. associated with polyagglutination, in vitro, including
To gain a better understanding of this patient’s outcome, Pneumococcus, Streptococcus, Staphylococcus, Clostridia,
they performed a prospective study to more closely Escherichia coli, Vibrio cholera, and influenza viruses.6
examine the relationship between NEC, T-activation, and Alterations in the RBC membrane resulting in poly-
hemolysis in patients in the neonatal intensive care unit agglutination can occur due to the action of enzymes or
(NICU). As stated in their introduction, “the causal rela- incomplete biosynthesis of RBC membrane components
tionship between T-activation and hemolysis has been or through inheritance of an uncommon haplotype.3
called into question by many clinicians and serologists.”1 The T-antigen is normally masked by a structure com-
Their work adds more information to this provocative posed of terminal N-acetyl neuraminic acid (NeuAc or
topic and also presents prospective data with more sialic acid), D-galactose, N-acetyl-D-galactosamine (Gal-
detailed laboratory testing than many previous studies. Nac), and serine or threonine. Upon the cleavage of the
The first observations of polyagglutination of RBCs terminal NeuAc by neuraminidase produced by microbial
date back to the 1920s when RBCs were unexpectedly agents, the T-antigen is uncovered, leaving behind D-
agglutinated by normal human adult sera that were galactose, GalNac, and serine or threonine.2 This antigen
ABO-compatible; this reaction did not occur with sera is carried on the M and N sialoglycoproteins. The exposed
from newborns.2,3 Hu € bener initially studied the phe- galactose residues, in the correct three-dimensional state,
nomenon of polyagglutination secondary to in vitro become the receptor for anti-T. Th activation appears to
contamination of blood samples with bacteria.4 Thom- represent an earlier and milder form of T-activation,
sen noticed that RBCs could change types from one day where less sialic has been cleaved from the RBC surface.
to the next and they could even be agglutinated with This type of activation is most common in newborns and
AB plasma, causing typing discrepancies. He called this their mothers.2,6,9 Tk activation is due to the action of
characteristic “panagglutinable” and theorized it was bacterial B-galactosidases rather than neuraminidase.3,9
secondary to RBCs agglutinating with bacteria.5 His stu- Tx activation has been described in children with pneu-
dent, Friedenreich, then showed that enzymes released mococcal infection. Tn activation is the result of a
from bacteria caused changes in RBC membranes that somatic mutation. In contrast with the other forms of T-
made them agglutinate. Due to the routine practice of activation, which are transient, Tn activation is persis-
using uncapped specimens that stood at room tempera- tent.6 When polyclonal reagents were in common use, T-
ture for many hours, these laboratories, performing activation could be identified through routine testing and
was usually first recognized by discrepancies in cell and
doi:10.1111/trf.14366 serum types. Now, with the use of monoclonal reagents,
C 2017 AABB
V other techniques have to be employed to seek out and
TRANSFUSION 2017;57;2553–2557 detect T-activation. The various forms of T-activation can

Volume 57, November 2017 TRANSFUSION 2553

 15372995, 2017, 11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/trf.14366 by Readcube (Labtiva Inc.), Wiley Online Library on [26/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EDITORIAL

be identified and distinguished from each other by treat- rare event in patients with T-activation and NEC, possi-
ment with various plant lectins. T-activation is character- bly due to the weakness of anti-T, exhibiting titers less
ized by reactivity with both the peanut lectin, Arachis than 64. Low titers of anti-T are in contrast with anti-A
hypogaea, and the soy lectin, Glycine soya. and anti-B in group O plasma, which are usually much
After 2 months of age, infants start to develop anti- higher in titer, but still only rarely cause hemolysis
T, reaching adult levels by the ages of 2 to 5 years old. when passively transfused. Likewise, Issitt and Anstee2
Anti-T is believed to develop due to exposure of bacte- found only weak titers of 2, 4, and 8 in tested plasma.
rial gut flora, similar to the appearance of the isoagglu- They likened anti-T to anti-P1 and Leb, antibodies that
tinins anti-A and anti-B. There was also speculation are not considered clinically significant.7,15,16 With sialic
that their formation is the result of childhood immuni- acid expression correlating with hemolysis of T-
zations.2 The appearance of anti-T is transient, lasting activated RBCs, there is no need for special blood prod-
weeks to months. Anti-T is an immunoglobulin (Ig)M ucts, except with profound desialylation and use in
antibody, most active at 48C (not usually active at 378C), plasma exchange.15
and does not fix complement. Based on these charac- Neuraminidase released by bacteria will break down
teristics, it is unlikely that anti-T can cause clinically sialic acid (NeuAc) in the RBC membrane, resulting in
significant hemolysis.2,3,6 Anti-T is not always present in damage that causes hemolysis. Early reports in animals
adults with T-activated RBCs, possibly due to neutrali- showed that T-activated RBCs had decreased survival and
zation by T-activated substances in the patient’s increased clearance in the presence of anti-T. To induce
plasma.2,8 T-activation, the RBCs were exposed to either neuramini-
The serologic properties of anti-T do not support its dase or influenza virus.7 This nonimmune mechanism of
role as a cause of clinically significant hemolysis. There hemolysis, in the setting of T-activation, can be explained
are a variety of nonimmunologic mechanisms that can by understanding the RBC structure. Sialic acid appears
cause RBC hemolysis, which may occur in infected to protect RBC membranes from degradation. When
patients. The direct effect of toxins, such as phospholi- sialic acid is removed by hydrolysis, there is increased
pase C (PLC), released from bacteria can damage the binding with A. hypogaea, showing that physically remov-
RBC membrane and lead to hemolysis. This effect is ing sialic acid, as opposed to immunologic actions, can
demonstrated in a report of an adult with sepsis second- destabilize the RBC membrane.17 Therefore, anti-T bind-
ary to Clostridia perfringens, E. coli, and enterococci, ing to the T-antigen is not causing the hemolysis, but it is
without evidence of T-activation, who died after massive the breakdown of sialic acid that causes RBC membrane
hemolysis. In this patient, PLC was found to increase in damage and subsequent hemolysis. The presence of sialic
concordance with levels of lactate dehydrogenase.10 acid is important to prevent complement-mediated lysis.
The pore-forming toxin perfringolysin O, a Sialic acid binds to complement factor H, enhancing its
cholesterol-dependent cytolysin, is more effective as cells affinity for C3b, inhibiting C3 convertase of the alternate
age, pointing toward a role in RBC destruction at the pathway. The enhancement of the classic pathway on cell
end of their life span. Likewise, pneumolysin, another surfaces by desialylation has been reported.15,18,19 Further
cholesterol-dependent cytolysin, can cause membrane evidence from animal models and models using RBCs
pore formation in RBCs leading to hemolysis. With anti- from healthy human blood donors show that there is
biotic use, as well as in the setting of sepsis, there can desialylation as RBCs age, which plays a role in RBC
be increases in susceptibility to cholesterol-dependent senescence. This pathway provides additional evidence of
cytolysin. When treatment with antimicrobials begins, the important role of sialic acid in maintaining RBC
there is an increase in the release of cytotoxins from lysis integrity.15,20 Therefore, anti-T-mediated hemolysis is
of bacteria that bind to membrane cholesterol that dependent on the degree of desialylation of RBCs, with-
increase the risk of RBC membrane lysis.11-13 Finally, out the presence of anti-T.15
there are case reports of drugs such as cephalosporins T-activation is most common in infants and young
causing acute hemolysis through their actions on children with bacterial infections, in particular in the
RBCs.14 setting of NEC with C. perfringens and nondiarrheal
Hemolysis by anti-T is usually weak and only sig- hemolytic uremic syndrome (HUS) with Streptococcus
nificant when there is profound desialylation of RBCs. pneumoniae. There are also case reports of T-activation
Anti-T hemolysin has been found to be weaker than in adults with bacterial infection.10,21
anti-A and anti-B hemolysins, speaking against its clini- It is important to recognize that hemolysis is more
cal significance. Using an in vitro system to assess anti- difficult to diagnose and assess in newborns and young
T-mediated hemolysis, Des Roziers and colleagues15 children. Since normal ranges for analytes are different
showed that hemolysis can significantly increase, inde- from adults, there should be caution in interpreting lab-
pendent of T antigen expression. These investigators oratory values. Sick newborns are subject to various
showed that hemolysis after plasma administration is a fluid shifts due to therapy and blood draws that can

2554 TRANSFUSION Volume 57, November 2017

 15372995, 2017, 11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/trf.14366 by Readcube (Labtiva Inc.), Wiley Online Library on [26/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EDITORIAL

lead to variability of hemoglobin levels.8 In addition, both HUS and invasive pneumococcal disease with
since pretransfusion testing of neonates is truncated to hemolytic anemia. They recommend that early testing
reduce iatrogenic blood loss, there are fewer opportuni- can be helpful to serve as a marker of invasive disease
ties to uncover T-activation. It is always important to once pneumococcus is diagnosed, since T-activation
formulate a broad differential diagnosis to rule out all appears to precede microangiopathic anemia and
other causes of hemolysis, when evaluating these very thrombocytopenia. Galectins, soluble proteins that have
ill, complex infants.7,8 an affinity for T-antigen, may also serve as an additional
There have been decades of case reports, case series, marker for T-activation.32
and discussions of the association between NEC in T-antigen is present not only on RBCs, but also on
infants, T-activation, hemolysis, and hemolysis after PLTs and renal glomerular cells, possibly having a role in
transfusion. Many cases of hemolysis reported in patients the thrombocytopenia and renal failure seen in HUS,
with NEC occurred in patients infected with although these findings may be solely related to the
C. perfringens. In this setting, neuraminidase from C. per- microangiopathic changes of HUS.3,7,34,35 There are case
fringens cleaves sialic acid from the RBC membrane, reports where children with severe HUS, who are only
exposing the T-antigen. In some patients who are T- transfused with washed blood products and undergo
activated and transfused with blood products from plasma exchange with albumin to avoid plasma, have
adults, whose plasma contains anti-T, severe and even good outcomes.35-38 As with NEC, there are conflicting
deadly hemolysis has been reported. In one study of four and incomplete data, creating controversy as to whether
patients, two children received standard transfusions or not washed components and low anti-T-titer plasma
before the diagnosis of NEC; one died and the other products improve outcomes; T-activation is associated
required two RBC exchanges, using washed RBCs sus- with hemolysis whether or not the patient has received a
pended in albumin. Two patients who were transfused transfusion; reports of no hemolysis after transfusion of
after the diagnosis and only received washed RBCs, anti-T-containing products; and reports of patients receiv-
washed platelets (PLTs), albumin, or protein fraction had ing only low-risk products with poor outcomes.4,7-9,34,39
no evidence of increased RBC destruction. Since this Waters and associates34 found that using fresh-frozen
publication7-9,22-25 and others cite worse outcomes due to plasma (FFP) or albumin for plasma exchange did not
T-activation and the temporal relationship with the trans- seem to have an impact on patient outcomes.
fusion and hemolysis, some have called for exclusive use Therefore, with decades of data, much of it conflict-
of washed and lower-risk blood products. Conversely, ing, where are we left? Can we make some decisions
some posit that T-activation is actually a marker of sever- and feel comfortable that these decisions will yield the
ity of disease, being associated with an increased risk for best patient outcomes? In the report by Moh-Klaren
the need for surgery, worse outcomes including higher and coworkers in this issue of TRANSFUSION, their
mortality, and the presence of C. perfringens and that the patient received antibiotics and FFP due to low
transfusion is not related to hemolysis.1,7,9,24,26-28 fibrinogen and then developed hemolysis, with tests
Although some publications show an association showing T-activation; the DAT and eluate were negative.
between T-activation and transfusion, there is no consis- Due to worsening clinical status, the patient underwent
tency in these reports. In some of the reports, serologic an exchange transfusion with washed RBCs suspended
data, such as testing to confirm presence of anti-T, the in 4% albumin. The infant then developed renal failure,
direct antiglobulin test (DAT), and temperature of testing disseminated intravascular coagulation, and multiple
may be missing or contradictory. The reported cases are organ failure leading to his death. Additional studies
not consistent; some patients with T-activation do not showed that the transfused FFP agglutinated
hemolyze, some without T-activation hemolyze, some neuraminidase-treated RBCs at 4 and 228C, but only
with T-activation hemolyze with unwashed products, and weakly at 378C. DTT treatment showed that the anti-T
some with T-activation hemolyze with washed products. was IgM and the titer at 48C was 8. No hemolysis of
It is important to remember that T-activation is rare and neuraminidase-treated RBCs was observed in vitro with
hemolysis with T-activation is rare.1,7-9,16,26,27,29-31 If there a serum sample from the plasma donor.1
is a true linkage between T-activation, transfusion, and Their accompanying prospective study included
hemolysis, this variation should not exist. 266 infants in the NICU who were tested for T-
Nondiarrheal HUS has also been associated with T- activation. Fifty-six of 238 patients without NEC were
activation, primarily in children infected with P. pneu- septic or were suspected of sepsis; only mild Th activa-
moniae.32 As with NEC, it is associated with a reduction tion was detected in one of these patients. Of 28
in sialic acid in the RBC membrane, secondary to patients with NEC, three had 41 reactivity with lectins.
neuraminidase, leading to hemolysis. Huang and col- They received unwashed RBCs and plasma-containing
leagues33 have shown that T-activation is a sensitive blood products without evidence of hemoglobinuria; all
marker for HUS in invasive pneumococcal disease, in patients were discharged from the hospital.1

Volume 57, November 2017 TRANSFUSION 2555

 15372995, 2017, 11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/trf.14366 by Readcube (Labtiva Inc.), Wiley Online Library on [26/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EDITORIAL

Based on these findings, Moh-Klaren and coworkers unnecessary medical interventions, understanding that
concluded that the poor outcome of their patient was doing more is not necessarily doing better. As transfu-
not due to passively transfused anti-T to their patient, sion medicine practitioners, we have been brought up
but was due to nonimmune destruction. Therefore, they with the precautionary principle that any risk needs to
concluded that there is no convincing evidence for rou- be mitigated, but based on literature that is more than
tine screening for T-activation in infants with NEC and 50 years old and more current studies, like that of Moh-
there is no reason for the use of plasma-containing Klaren and colleagues, we have not proven causality. It
blood products to be withheld from infants with NEC, may be time to consider that testing to determine the
even in the presence of T-activation. The strict avoid- need for special blood products, and supplying special
ance of plasma could be harmful, especially in the set- blood products for the treatment of infants and children
ting of coagulopathy, and manipulations of cellular with T-activation is not necessary.
products and plasma could create potentially dangerous
delays.1
CONFLICT OF INTEREST
I am in agreement with Moh-Klaren and coworkers.
There are many robust mechanisms that account for The author has disclosed no conflicts of interest.
the nonimmune RBC destruction. In addition, patients
studied have infections with exposure to bacterial toxins Susan D. Roseff
and are being treated with antibiotics that can play a Department of Pathology
role in increased RBC destruction. Reports of children VCU School of Medicine
with the highest risk of T-activation and hemolysis Richmond, VA
show that those children are usually the sickest, with e-mail: [email protected]
multiple comorbidities. Some patients might have
needed early transfusion before T-activation was identi-
REFERENCES
fied, due to the severity of their disease, explaining
poorer outcomes in transfused patients.7 I believe that 1. Moh-Klaren J, Bodivit G, Jugie M, et al. Severe hemolysis
T-activation may represent a correlate of the degree of following plasma transfusion in a neonate with necrotiz-
toxin present and serve as a possible marker for disease ing enterocolitis, Clostridium perfringens infection and
severity. This information could be helpful for patients red-cell T-polyagglutination. Transfusion 2017;57:2571-77.
and their parents to understand when discussing possi- 2. Issitt PD, Anstee DM, editors. Applied blood group
ble prognosis and outcomes. serology. 4th ed. Durham (NC): Montgomery Scientific
It is difficult to find fault with practitioners trying Publications; 1998. p. 1097-105.
to follow a conservative course of treatment for sick 3. Judd WJ. Review: polyagglutination. Immunohematology
children, fragile newborns and premature infants by 1992;8:58-69.
providing special blood products. On the other hand, 4. Judd WJ. Microbial associated forms of polyagglutination
there are no consistent data that transfusion of blood (T, Tk and acquired-B). In Beck ML, Judd WJ, editors. Poly-
products with anti-T is the etiology for hemolytic agglutination: a technical workshop. 1st ed. Washington,
events. The serologic principles we adhere to show that DC: AABB 1980:23-53.
anti-T is not clinically significant and titers are usually 5. Pierce SR, Reid ME. Bloody brilliant: a history of blood
low; there is also no proven “safe” threshold. Since anti- groups and blood groupers. Bethesda (MD): AABB; 2016.
T begins to appear in infants after the age of 2 months p. 248-9, 393-3, 449.
and those older than 2 years may already have adult 6. Mollison PL, Engelfriet CP, Contreras, M. Blood transfusion
levels that do not appear to be causing hemolysis, it in clinical medicine. 10th ed. London: Blackwell Science;
seems unlikely that passive anti-T is dangerous. 1997. p. 233-8.
Striving to work toward evidence-based treatments 7. Crookston KP, Reiner AP, Cooper LJ, et al. RBC T activation
that result in the best patient outcomes, I think we can and hemolysis: implications for pediatric transfusion man-
conclude that there is not good evidence for worsening agement. Transfusion 2000;40:801-12.
outcomes being directly caused by transfusing standard 8. Ramasethu J, Luban N. T activation. Br J Haematol 2001;
blood products in the setting of T-activation in patients 112:259-63.
with NEC and HUS. Due to the rarity of hemolysis with 9. Eder AF, Manno CS. Does red-cell T activation matter? Br J
T-activation and reports from various colleagues of the Haematol 2001;114:25-30.
inability to consistently have commercial reagents read- € bl W, Mostbeck B, Hartleb H, et al. Investigation of the
10. Hu
ily available, conducting a randomized controlled trial pathogenesis of massive hemolysis in a case of Clostrid-
may not be feasible. We are in an era of changes in ium perfringens septicemia. Ann Hematol 1993;67:145-7.
the paradigms of medical practice. There are many 11. Bokori-Brown M, Petrov PG, Khafaji MA, et al. Red blood
efforts to reduce unnecessary laboratory testing and cell susceptibility to pneumolysin: correlation with

2556 TRANSFUSION Volume 57, November 2017

 15372995, 2017, 11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/trf.14366 by Readcube (Labtiva Inc.), Wiley Online Library on [26/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
EDITORIAL

membrane biochemical and physical properties. J Biol 26. Williams RA, Brown EF, Hurst D, et al. Transfusion of
Chem 2016;291:10210-27. infants with activation of erythrocyte T antigen. J Pediatr
12. Anderson R, Tintinger G, Cockeran R, et al. Beneficial and 1989;115:949-53.
harmful interactions of antibiotics with microbial patho- 27. Osborn DA, Lui K, Pussell P, et al. T and Tk antigen activa-
gens and the host innate immune system. Pharmaceuti- tion in necrotising enterocolitis: manifestations, severity of
cals 2010;3:1694-1710. illness, and effectiveness of testing. Arch Dis Child Fetal
13. Iwamoto M, Nakamura M, Mitsui K, et al. Membrane dis- Neonatal Ed 1999;80:F192-7.
organization induced by perfringolysin O (theta-toxin) of 28. Kirsten GF, Smith J, Pieper C, et al. The necessity for T-
Clostridium perfringens—effect of toxin binding and self- cryptantigen activation screening in babies with nerotising
assembly on liposomes. Biochim Biophys Acta 1993;1153: enterocolitis. S Afr Med J 1996;85:546-8.
89-96. 29. Wang LY, Chan YS, Chang FC, et al. Thomsen-Friedenreich
14. Shulman IA, Arndt PA, McGehee W, et al. Cefotaxime- activation in infants with necrotizing enterocolitis in Tai-
induced immune hemolytic anemia due to antibodies wan. Transfusion 2011;51:1972-6.
reacting in vitro by more than one mechanism. Transfu- 30. Heddle NM, Blajchman MA, Bodner N, Sacher R. Absence
sion 1990;30:263-6. of hemolysis of T-activated red cells following infusion of
15. Des Roziers NB, Bodivit G, Chadebech P, et al. Anti-T hae- blood products [abstract]. Transfusion 1977;17:686.
molysins: the effects of sialic acid removal and 2- 31. Engelfriet CP, Reesink HW, Strauss RG, et al. Blood transfu-
aminoethylisothiouronium bromide treatment of erythro- sion in premature or young infants with polyagglutination
cytes on immune lysis. Vox Sang 2011;100:401-8. and activation of the T antigen. Vox Sang 1999;76:128-32.
16. Boralessa H, Modi N, Cockburn H, et al. RBC T activation 32. Des Roziers NB, Chadebech P, Bodivit G, et al. Red blood
and hemolysis in a neonatal intensive care population: cell Thomsen-Friedenreich antigen expression and
implications for transfusion practice. Transfusion 2002;42: galectin-3 plasma concentrations in Streptococcus
1428-34. pneumoniae-associated hemolytic uremic syndrome and
17. Springer GF, Desai PR. Extent of desialation of blood hemolytic anemia. Transfusion 2015;55:1563-71.
group MM, NN, and MN antigens required for reactivity 33. Huang DT, Chi H, Lee HC, et al. T-antigen activation for
with human anti-T antibody and Arachis hypogaea lectin. prediction of pneumococcus-induced hemolytic uremic
J Biol Chem 1982;257:2744-6. syndrome and hemolytic anemia. Pediatr Infect Dis J
18. Liu C, Marshall P, Schreibman I, et al. Interaction between 2006;25:608-10.
terminal complement proteins C5b-7 and anionic phos- 34. Waters AM, Kerecuk L, Luk D, et al. Hemolytic uremic
pholipids. Blood 1999;93:2297-2301. syndrome associated with invasive pneumococcal disease:
19. Marshall P, Hasegawa A, Davidson EA. Interaction between the United Kingdom experience. J Pediatr 2007;151:140-4.
complement proteins C5b-7 and erythrocyte membrane 35. Hopkins CK, Yuan S, Lu Q, et al. A severe case of atypical
sialic acid. J Exp Med 1996;184:1225-32. hemolytic uremic syndrome associated with pneumococ-
20. Durocher JR, Payne RC, Conrad ME. Role of sialic acid in cal infection and T activation treated successfully with
erythrocyte survival. Blood 1975;45:11-20. plasma exchange. Transfusion 2008;48:2448-52.
21. Judd WJ, Oberman HA, Flynn S. Fatal intravascular hemo- 36. Petras ML, Dunbar NM, Filiano JJ, et al. Therapeutic
lysis associated with T-polyagglutination. Transfusion plasma exchange in Streptococcus pneumoniae-associated
1982;22:345-6. hemolytic uremic syndrome: a case report. J Clin Apher
22. Novak RW. Bacterial-induced RBC alterations complicating 2012;27:212-4.
necrotizing enterocolitis. Am J Dis Child 1984;183-5. 37. Cochran JB, Panzarino VM, Maes LY, et al. Pneumococcus-
23. Klein RL, Novak RW, Novak PE. T-cryptantigen exposure in induced T-antigen activation in hemolytic uremic syn-
neonatal necrotizing enterocolitis. J Pediatr Surg 1986;21: drome and anemia. Pediatr Nephrol 2004;19:317-21.
1155-8. 38. Erickson LC, Smith WS, Biswas AK, et al. Streptococcus
24. Placzek MM, Gorst DW. T activation haemolysis and death pneumonia-induced hemolytic uremic syndrome: a case
after blood transfusion. Arch Dis Child 1987;62:743-4. for early diagnosis. Pediatr Nephrol 1994;8:211-3.
25. Seger R, Joller P, Kenny A, et al. Potential hazards of 39. Oliver JW, Akins RS, Bibens MK, et al. Pneumococcal
blood-transfusion in Clostridia-associated necrotizing induced T-activation with resultant thrombotic microangi-
enterocolitis. Lancet 1979;1:48-9. opathy. Clin Med Insights Pathol 2010;3:13-7.

Volume 57, November 2017 TRANSFUSION 2557